Modelling the fast dynamics of power converters is of growing concern in power grids and Microgrids. Dynamic phasor (DP) concept has been widely applied to switched power converter for modelling fast transients efficiently due to the inherent frequency shift property of DPs. The dynamics introduced by the DC/AC power converters depend on the controllers, which are implemented either in the stationary frame or on the synchronous reference frame (SRF). Hybrid closed-loop modelling methods that consider DP-based power converter plant model are not established and the applicability and accuracy of such hybrid approaches are not fully understood. This paper attempts to address this gap by proposing two hybrid closed-loop modelling approaches: Hybrid DP-DQ and Hybrid DP-EMT and discusses the applicability and accuracy for various controller types. Furthermore, this paper presents a DP switched model of single and three phase two level power converter and discusses the selection of harmonics to reduce model complexity. The proposed hybrid approaches were validated against detailed switched power converter models and for a wide range of scenarios, the Hybrid DP-EMT method is found to be superior compared other methods. Finally, application dependent recommendations are made for the selection of suitable hybrid closed-loop model for the accurate simulation of single phase and three phase power converters.

This work develops and benchmarks surrogate models for Dynamic Phasor (DP) simulation of electrical drives. DP simulations of complex systems may be time-consuming due to the increased number of equations. Thus, it is desirable to have a datadriven approach to compute the critical state/control variables and power losses. The surrogate models are intended to be used as a steady-state equivalent of the DP simulation model. We consider the Gaussian Process (GP), Multi Layer Perceptron, and Random Forest as surrogate models. Among other techniques, GPs are found to have good accuracy. Moreover, GPs are dataefficient and have desirable properties, such as built-in uncertainty quantification. The study shows that the GP performs better compared to other techniques in terms of the Mean Squared Error of the prediction, while still being very fast to evaluate. We illustrate the potential of these surrogate models to also predict transient behavior.